Benchmarking Self-Supervised Learning on Diverse Pathology Datasets
This work addresses the annotation bottleneck in computational pathology by benchmarking and adapting SSL methods, offering incremental improvements for domain-specific applications.
The authors conducted a large-scale study comparing self-supervised learning methods on diverse pathology datasets, finding that domain-aligned pre-training consistently outperforms ImageNet pre-training, with improvements in linear evaluations, fine-tuning, and low-label regimes, and applied SSL to nuclei instance segmentation for the first time, showing large performance gains.
Computational pathology can lead to saving human lives, but models are annotation hungry and pathology images are notoriously expensive to annotate. Self-supervised learning has shown to be an effective method for utilizing unlabeled data, and its application to pathology could greatly benefit its downstream tasks. Yet, there are no principled studies that compare SSL methods and discuss how to adapt them for pathology. To address this need, we execute the largest-scale study of SSL pre-training on pathology image data, to date. Our study is conducted using 4 representative SSL methods on diverse downstream tasks. We establish that large-scale domain-aligned pre-training in pathology consistently out-performs ImageNet pre-training in standard SSL settings such as linear and fine-tuning evaluations, as well as in low-label regimes. Moreover, we propose a set of domain-specific techniques that we experimentally show leads to a performance boost. Lastly, for the first time, we apply SSL to the challenging task of nuclei instance segmentation and show large and consistent performance improvements under diverse settings.